Homebrew-core: Formulae without Sierra bottles

[zmwangx]

Jenkins finally managed to chew through the daunting Sierra build queue. Below are the remaining ones that defied (not necessarily only) Sierra for one reason or another.

The list is generated with the script https://gist.github.com/zmwangx/89f07670c8e42f58434bd6d15653ecf6. You can also pass a blacklist to the script to get a more workable list (e.g., putting qt and qt5 in the blacklist blocks the entire Qt tree). See need-bottle -h.

---

Formulae that made the top 5000 install events chart during September 2016, ordered by popularity:

• [x] qt5
• [x] xctool (https://github.com/facebook/xctool/issues/716)
• [ ] swiftlint (swift 2.2, https://github.com/realm/SwiftLint/issues/815)
• [x] pypy (upstream issue, #5689)
• [ ] pyqt (qt, #1705)
• [x] pyqt5
• [x] qscintilla2 (https://github.com/Homebrew/homebrew-core/pull/5786)
• [ ] swiftgen (swift 2.2, https://github.com/AliSoftware/SwiftGen/issues/169)
• [x] qwt
• [x] allegro (#6017)
• [x] pyside
• [x] shiboken
• [x] gnuradio
• [x] upx (#6264, https://sourceforge.net/p/upx/bugs/248/, https://github.com/upx/upx/issues/4)
• [ ] infer (need to wait for 0.9.4 according to #5130)
• [x] jack (#5487)
• [x] qca
• [x] qwtpolar
• [x] binwalk
• [x] git-cola (#6210)
• [x] folly (#5548)
• [x] gpsbabel
• [x] qcachegrind
• [x] radare2
• [x] libewf (libyal/libewf#60, #5737)
• [x] dvm (#5587)
• [x] irrlicht (#5592)
• [x] tcptrace (#6223)
• [ ] imagesnap (project dead since 2011, but 10.11 bottle works on 10.12 unmodified)
• [x] uwsgi (https://github.com/Homebrew/homebrew-core/pull/5528)
• [x] log4cxx
• [ ] capnp (#5773, https://github.com/sandstorm-io/capnproto/issues/375)
• [ ] swift (need to upgrade to 3.0)
• [ ] fb-adb (has newer release, which unfortunately requires dupes/make)
• [x] fpc (#5732)
• [ ] ctorrent (redefinition of clockid_t)
• [x] syncthing-inotify
• [ ] pgloader (has newer release, updating resources seems to be a royal pain though)
• [ ] volatility (Pillow problem, complexity of homebrew/python/pillow isn't confidence-inspiring; I wonder if Pillow is really needed, not immediately obvious)
• [ ] libqglviewer (qt, #1705)
• [x] sift (https://github.com/svent/sift/pull/80)
• [x] shairport-sync
• [x] pyside-tools
• [ ] otto (go 1.7+:-X flag requires argument of the form importpath.name)
• [x] ace
• [ ] ircii (#6404)
• [ ] juju (#5911)
• [x] dcmtk
• [ ] io (seems good to go)
• [ ] xml2 (#6405)
• [ ] tomcat-native (apr problem)
• [x] moc
• [x] natalie
• [ ] open-zwave (broken dylib link in build path)
• [ ] rack (https://github.com/rackspace/rack/issues/446, go 1.7+:-X flag requires argument of the form importpath.name)
• [x] codequery
• [x] zurl
• [ ] libgroove (redefinition of clockid_t and clock_gettime)
• [x] opencsg
• [x] kube-aws
• [x] libmongoclient
• [ ] shibboleth-sp
• [ ] bashdb (#6407)
• [x] bokken
• [x] castxml
• [ ] vice (references QuickTime/QuickTime.h)
• [x] debianutils
• [ ] taylor (swift 2.2)
• [ ] distcc (migrated from Google Code to https://github.com/distcc/distcc, but no new release)
• [x] darkice
• [x] pushpin
• [x] gsoap
• [ ] dc3dd (seems good to go)
• [x] sqliteman
• [x] aurora-cli (#6006)
• [x] djview4 (https://github.com/Homebrew/homebrew-core/pull/5736)
• [ ] gecode (qt5, maybe good to go)
• [ ] libfaketime (clock issue, code assumes CLOCK_REALTIME undefined on macOS, masking #include <mach/clock.h>)
• [x] frescobaldi
• [ ] libbtbb (#6401)
• [x] qjackctl
• [ ] chuck (appear to be having trouble linking /System/Library/PrivateFrameworks/MultitouchSupport.framework properly)
• [ ] dshb (https://github.com/beltex/dshb/issues/36)
• [x] puddletag
• [ ] libbladerf (redefinition of clockid_t)
• [ ] qjson (qt, #1705)
• [x] ssdb (ideawu/ssdb#995, #5734)
• [x] dspdfviewer
• [x] st
• [ ] coin (qt, #1705)
• [x] mftrace
• [ ] darkstat (redefinition of clockid_t)
• [ ] makepkg (#6402)
• [ ] aws-apigateway-importer (#6403)
• [x] cfengine (https://github.com/Homebrew/homebrew-core/pull/5513)
• [ ] fstar (those ocaml resources are causing conflicts and a lot of pain, and ocaml install can choose to ignore your cached downloads anyway; should just admit resources aren't good for everything, and let the package manager do the heavylifting.)

---

Obscure formulae that didn't make the top 5000 install events chart (some of these need to be thrown into boneyard — no mercy for the weak):

• [ ] activemq-cpp
• [ ] aiccu
• [x] automoc4
• [ ] blucat
• [ ] buildapp
• [ ] bzr-explorer
• [ ] coccinelle
• [ ] compface
• [x] ctail
• [x] cuty_capt
• [x] dirt
• [ ] dsd
• [ ] exim
• [ ] ezlupdate
• [x] fastd (upstream issue, #6012)
• [ ] fdroidserver
• [x] gammaray
• [ ] globus-toolkit
• [ ] intercal
• [ ] ipbt
• [ ] ironcli
• [ ] jxrlib
• [ ] libechonest
• [ ] libetonyek
• [ ] libfabric
• [ ] liblastfm
• [ ] libodfgen
• [x] ltc-tools
• [x] md
• [ ] mdds
• [ ] mighttpd2
• [ ] mimms
• [ ] mmix
• [ ] nordugrid-arc
• [ ] openvdb
• [ ] parrot
• [x] qbs
• [ ] qpm
• [ ] qtplay
• [ ] quazip
• [ ] qxmpp
• [x] rcssserver
• [ ] rdate
• [x] remctl
• [ ] saldl
• [ ] schismtracker
• [ ] sfk
• [ ] snappystream
• [ ] srclib
• [ ] swish-e
• [ ] tbox
• [ ] tevent
• [ ] tlassemble
• [ ] uberftp
• [ ] ubertooth
• [x] valkyrie
• [x] weboob
• [ ] writerperfect
• [ ] wyrd
• [ ] xar-mackyle

[zmwangx]

I've ordered the list by popularity so hopefully the priority is clearer now.

[leee]

@zmwangx - do you intend on updating this list(s) as Formula items get updated? (c.f. #5776)

[zmwangx]

@leee Maintainers should check off items directly.

For non-maintainers: if you notice something that should be checked off but hasn't been, please leave a comment, and I (or another maintainer) will get to it. Thanks.

[Moisan]

bokken depends on radare2 which is now fixed. It seems to build correctly now.

[Moisan]

gsoap and dcmtk are fixed.

[zmwangx]

@ilovezfs I've added a reason or potential solution for every reasonably popular formula (just can't get excited about the rest of them).

dc3dd and io seem to be compiling fine now, should try them out on Sierra Testing; gecode is a Qt5 revdep, maybe good to go, not sure because I don't want to install Qt.

A bunch of clock-redefinition should be easy to fix, will leave that to you so that the replaced gibberish is consistent across the code base.

There might be other low hanging fruits which I forgot to mention.

[ilovezfs]

@zmwangx #6408

[ilovezfs]

Anyone who has been following or contributing to this issue, please use #6408 from now on.